The Arbitrarian is the work of David Sparks, a graduate student in Political Science at Duke University. I’m interested in politics, statistics, information design, sports, and everything else. Feel free to visit my personal homepage at Duke, or see a list of other work I’ve produced.
15 responses so far ↓
Ty // April 27, 2008 at 10:17 am |
Although I don’t really understand half of what I have seen from you , it is quickly evident that you are a bright guy. I coach a girls varsity basketball team and I was purusing the web to try and find a decent stat comparison spreadsheet to fairly evaluate and grade my players overall contribution to the team. I stumbled upon your site. Good stuff.
Neal // May 13, 2008 at 9:19 am |
David, I’m very interested in the People’s Statistic Project. I fool around with stats myself, but not the extent that you and other APBR folks do so. However, when I first looked at the PS results, it occurred to me that it looked more like a high-concept popularity contest than an actual “contributes to wins” stat. From a marketing point of view, I think that might actually have some value. You mentioned in a forum that many of the surveyed seem to be trying to tweak the results to make MJ the #1 all-time. I think if you included dunks in the survey that might put him over the top, and yield a weighting system that could predict a players “star-power” in the modern age. Chris Paul is becoming very popular, but can you imagine if he dunked on guys?
Chip Crain // May 14, 2008 at 9:23 am |
Just thought you would like to know I have posted your interview. It can be found at http://3shadesofblue.blogspot.com/2008/05/arbitrary-interview-with-arbitrarian.html
khandor // July 31, 2008 at 10:44 am |
David,
The contribution you just had published on Hardwood Paroxysm … is OUTSTANDING.
Bravo!
A Tour de Force.
khandor // July 31, 2008 at 11:14 am |
Secondly.
Given the respective weights you’ve mentioned, thus far, am I reading your work correctly … if I assert from this initial material that the two measures which receive the most weight in your metric are:
* or 1.0192830
* ‘total rebounds’ [or/1.0192830 + dr/ 0.5173558]
[considering the number of 'rebounds' there are in a game, relative to the number of 'other events' which happen]
d sparks // July 31, 2008 at 12:08 pm |
khandor: Interesting idea, to look at the stats from which the most points were gained or lost. Here’s the breakdown, summing statistics since 1979-80:
pts: 6384067
fgx: -2858370
ftx: -240559.2
as: 1631794
or: 839599.7
dr: 953948.7
st: 828238.8
bk: 312690.8
to: -1401794
pf: -320387.7
So, the biggest source has been scoring (somewhat obviously), and the biggest drain has been missed fgas. It appears that defensive rebounds have contributed more than offensive boards, due to the much greater number of drs than ors.
khandor // July 31, 2008 at 1:13 pm |
David,
Had an interesting exchange with Kevin Broom a while back, re: the relationship between ‘rebounds garnered in a specific game’ and ‘winning that game’ based on ‘regression analyses’ of traditional boxscore information and his adherence to the ‘4 Factors’ theories of Dean Oliver.
Mentioned to him that some years ago (actually in the mid ’90’s), as best I could recall, the research studies I read indicated that the most SIGNIFICANT relationship existed between ‘rebounds garnered in a specific game’ and ‘winning that specific game’ when using regression analysis of traditional boxscore stats (not including actual Pts Scored, of course, as you’ve identified above).
Kevin indicated that, according to the work he’d seen, my perception was inconsistent with the information he had at-hand (i.e. that of Dean Oliver and others).
Based on your specific metric and your use of regression analysis would you be able to corroberate my perception, in this regard, or not?
khandor // August 4, 2008 at 11:03 am |
David,
What happened to your article, “Anything is Possible” in your ‘Top Posts’ section?
It was there on last week … but now it’s disappeared.
d sparks // August 4, 2008 at 12:32 pm |
Khandor: I’m not sure why it dropped off the Top Posts lists, but you can find it here: http://arbitrarian.wordpress.com/2008/06/18/anythings-possible/
Sorry it’s taking me so long to follow up on your question about rebounding… between writing this week’s HP post and Real Life, I’m pretty busy.
khandor // August 4, 2008 at 1:20 pm |
Hi, David.
Thanks a bunch.
No problem, from my end … Me, too, in fact.
Too much to handle, all at once.
ASAP … will be just fine.
khandor // August 7, 2008 at 3:47 pm |
Solid job, once again, with your latest contribution over at HP.
When you get a chance, please remember my question in this section.
Thanks
khandor // August 20, 2008 at 2:44 pm |
Hi, David.
Still no answer worthcoming?
[just a friendly reminder
]
khandor // September 6, 2008 at 6:04 am |
Hi, David.
Still no answer forthcoming?
khandor // October 2, 2008 at 5:56 am |
Hi, David.
Is there going to be an answer provided for the question I asked?
Hopefully there is.
khandor // November 4, 2008 at 9:42 am |
Hi, David.
I’m still waiting for an answer to my specific question.